- xserver-1.3.0-randr-preferred-mode-fix.patch: Fix infinite loop on X
startup when a mode is requested in the config file. (#318731)
- Fix License tag.
- xserver-1.3.0-accidental-abi.patch: Make sure some symbols from parser/
get exported, since apparently the intel driver uses them despite their
not being in the documented ABI list. Thanks guys.
- xserver-1.3.0-randr-updates.patch: Default ModeDebug to TRUE, better to
have too much information than too little.
- xorg-x11-server-1.0.1-fpic-libxf86config.patch: Build the parser library
with hidden symbols to shrink pyf86config a bit.
- xserver-1.3.0-intel-by-default.patch: Use intel, not i810, when starting
without a config file.
- Enable maintainer mode when building so I swear at autotools less.
- xserver-1.3.0-randr-updates.patch: Backport randr from server git
This contains a lot of fixes since 1.3.0 went out, and saves
us backporting each fix individually
- xserver-1.3.0-less-randr-fakerama.patch - dropped
- xorg-x11-server-1.2.0-maxpixclock-option.patch - dropped
- xserver-1.3.0-edid-quirk-backports.patch - dropped
- xserver-1.2.0-honor-displaysize.patch - dropped
- xserver-1.3.0-honor-displaysize.patch - reapply to new code layout
- xserver-1.3.0-randr12-config-hack.patch - rebased
- xserver-1.3.0-mesa7.patch: Add support for building against mesa 7.0.1
along with DRI zero-copy TFP hopefully
- xserver-1.3.0-exaupgrade.patch: Add updated EXA support
- dropped xserver-1.2.99.901-xephyr-crash-at-exit.patch - upstream
- rebase xorg-x11-server-1.1.1-offscreen-pixmaps.patch to xserver-1.3.0-newglx-offscreen-pixmaps.patch
- dropped xorg-x11-server-1.1.1-glcore-visual-matching.patch - fixed upstream
- xserver-1.3.0-composite-version.patch: Force the server to report the
Composite extension version it supports, not simply the version defined
by the protocol headers it was built against.
- xserver-1.3.0-randrama-no-zero-screens.patch: For RANDR 1.2's fake
Xinerama info, don't report Xinerama as being active if there are no
RANDR 1.2 CRTCs active for that screen. (#234567)
- xserver-1.3.0-arm-iopl.patch: Add __arm__ conditionals to many #ifdefs.
- xserver-1.3.0-randrama-no-zero-screens.patch: For RANDR 1.2's fake
Xinerama info, don't report Xinerama as being active if there are no
RANDR 1.2 CRTCs active for that screen. (#234567)
- xserver-1.3.0-arm-iopl.patch: Add __arm__ conditionals to many #ifdefs.
- xorg-x11-server-1.1.1-offscreen-pixmaps.patch: Fix a crash when activating
GLX_EXT_texture_from_pixmap without XAA.
- xserver-1.3.0-randr12-config-hack.patch: If a Modes line is given in
the Screen section, and no PreferredMode option is given for a RANDR 1.2
monitor, use the first mode in the Modes line as the preferred mode.
Fixes anaconda ugliness on monitors larger than 800x600. (#238991)
- xserver-1.3.0-xkb-and-loathing.patch: Ignore (not just block) SIGALRM
around calls to Popen()/Pclose(). Fixes a hang in openoffice when
opening menus.
- Modify BuildRequires to use the virtual protocol Provides.
- xserver-1.3.0-no-prerelease-warning.patch: Hush the useless prerelease
warning if we happen to be building one (and even if not).
- xserver-1.3.0-pci-device-enable.patch: Make sure the PCI device is enabled
in sysfs before we start touching it, otherwise, armageddon.
- xserver-1.3.0-pci-bus-count.patch: Allocate the PCI bus array dynamically,
so as not to run off the end of it.
- xserver-1.3.0-mmap-failure-check.patch: Check for failure when mmap'ing
bus memory. (#234073)
- xserver-1.3.0-rom-search.patch: Look for the sysfs ROM file in the (flat)
/sys/bus/pci/devices tree, instead of the (bus-topological) /sys/devices,
so we don't fail to find ROMs merely because they're behind a bridge.
- xserver-1.3.0-no-pseudocolor-composite.patch: Refuse to initialize
Composite when Render is missing or when the root window is using
a pseudocolor visual. (#217388)
- xserver-1.3.0-xnest-exposures.patch: Fix Motif app redraw in Xnest. (#229350)
- Various backports from git master:
- xserver-1.2.0-xfixes-clientgone-check.patch: Avoids a crash when sending
events to clients that just disconnected.
- xserver-1.2.0-os-memory-leak.patch: Plugs a per-connection memory leak.
- xserver-1.2.0-int10-rdtsc.patch: Implement rdtsc in the int10 emulator.
- xserver-1.2.0-glcore-visual-count.patch: Count glcore visuals properly,
fixes crash at exit.
- xorg-x11-server-Red-Hat-extramodes.patch:
- Add 1360x768 normal and reduced-blanking.
- Add reduced-blanking versions of 1680x1050 and 1920x{1200,1080}.
- Remove the >60Hz versions of 2560x1600. Even leaving the 60Hz timing is
kind of ridiculous, since every real LCD that size I've seen uses the
reduced blanking timings. But presumably if you have that nice of a
monitor, you also have a video card with working DDC.
- Massive spec formatting and style cleanup.
- Build Xdmx on all arches.
- Enable GL support even on non-DRI machines.
- Re-add DRI to ppc64.
- Update BuildRequires to current versions.
- Remove some bogus Requires.
- RHEL5 sync:
- xorg-x11-server-1.1.1-maxpixclock-option.patch: Allow the maximum pixel
clock of a monitor to be specified in the config file.
- xorg-x11-server-1.1.1-glcore-visual-matching.patch: Fix a client crash
when creating software indirect GLX contexts.
- xorg-x11-server-1.1.1-vt-activate-is-a-terrible-api.patch: During server
init, abort if either VT activation ioctl fails. During shutdown, be
sure to wait for the VT switch to finish before exiting.
- xorg-x11-server-1.1.1-no-scanpci.patch: Drop scanpci, it's huge and
there's no added value relative to lspci.
- xorg-x11-server-1.1.1-spurious-libxf1bpp-link.patch: Minor linktime
fixup. There's no reason for libxf4bpp to link against libxf1bpp.
- Do not try own /usr/lib/pkgconfig in sdk package.
- Drop dependency on xorg-x11-fonts-base now that we compile in
'fixed' and 'cursor' fonts.
- Drop xorg-redhat-die-ugly-pattern-die-die-die.patch; use -br options
instead.
- xorg-x11-server-1.1.1-vt-activate-is-a-terrible-api.patch: Since the
VT_ACTIVATE/VT_WAITACTIVE pair are never guaranteed to successfully
complete, set a 5 second timeout on the WAITACTIVE, and retry the pair
until we win. (#207746)
- xorg-x11-server-1.1.0-pci-scan-fixes.patch: Partial revert to unbreak some
(but not all) domainful machines, including Pegasos. (#207659)
- xorg-x11-server-1.1.1-pclose-confusion.patch: Be sure to call Pclose()
on pipes created with Popen(), since the additional magic done by Popen()
relative to popen() is not undone by plain pclose(). (Third base!)
- xorg-x11-server-1.1.1-edid-hex-dump.patch: Backport EDID hex dump code
from git.
- xorg-x11-server-1.1.1-always-mouse-thyself.patch: Fix the check to look
for mouse/void drivers in the running layout, as opposed to the config file,
so as not to synthesize two mouse devices.
- xorg-x11-server-1.1.1-infer-virtual.patch: Be slightly more paranoid about
setting line pitch, and rescan the mode list after pruning to re-validate
the estimated virtual size.
- Update xorg-x11-server-1.1.1-offscreen-pixmaps.patch to log transitions.
- Update xorg-x11-server-1.1.0-tfp-damage.patch to always bind to
GL_TEXTURE_RECTANGLE_ARB target.
- Add xorg-x11-server-1.1.1-aiglx-happy-vt-switch.patch to fix VT
switching (and suspend/resume) when using AIGLX. (#199692, fdo #7916).
- Bump mesa source and libGL BuildRequires.
- Update mesa-6.5.1 patch to work with 6.5.1 rc1 (slang_version_syn.h
renamed to slang_pp_version_syn.h).
- xorg-x11-server-1.1.1-infer-virtual.patch: Only flag modes as preferred
if the EDID block says to.
- xorg-x11-server-1.1.1-mode-sort-kung-fu.patch: Enforce a sort order on
modes during lookup: builtin before driver before userdef before other,
and preferred modes within a class before others in that class.
- xorg-x11-server-1.1.1-xvfb-composite-crash.patch: Fix Xvfb's -render flag
to also disable the Composite extension.
- xorg-x11-server-1.1.1-mesa-6.5.1.patch: Update build system to account for
Mesa 6.5.1 snapshots.
- xorg-x11-server-1.1.0-edid-mode-injection-2.patch: Add all available
standard timings from EDID rather than just the first five.
- xorg-x11-server-1.1.1-infer-virtual.patch: When no modes or virtual size
are given in the config file, attempt to pick a sensible one by examining
the EDID modes and physical geometry. Also generally make the server
aware of driver-provided modes.
- xorg-x11-server-1.1.1-edid-quirks-list.patch: Redo, since the property I was
checking for is both fairly common and fairly predictable.
- xorg-x11-server-1.1.1-always-mouse-thyself.patch: If we lack a mouse
device in the config, and the user hasn't asked for any void devices,
synthesize a mouse section. (#200347)
- xorg-x11-server-1.1.1-edid-quirks-list.patch: Better formatting.
- xorg-x11-server-1.1.1-builderstring.patch: Enable the builder info
string at configure time;
- ... and use it to print the package name and version.
- xorg-x11-server-1.1.1-defaultdepth-24.patch: Set default depth to 24.
- xorg-x11-server-1.1.1-edid-quirks-list.patch: Add EDID quirks list as
an experiment; needs a better solution though.
- Comment out the 848x480 modes from the extramodes patch. Any panel that
wants it should be doing EDID injection by now, and it screws up
autoconfig by _just_ fitting in the ranges for 800x600.
- xorg-x11-server-1.1.1-selinux-awareness.patch: Added for new Mesa
selinux code.
- xorg-x11-server-1.1.1-Xdmx-render-fix-fdo7482.patch: Backport a Render
fix for Xdmx.
- xorg-x11-server-1.1.1-no-composite-in-xnest.patch: Disable Composite in
Xnest, as it's known not to work.
- Fix default font path to match the config file we used to generate.
- Fix default module set to match the config file we used to generate.
- Disable use of TLS GLX dispatch to match Mesa selinux nonsense.
- xorg-x11-server-1.1.1-selinux-awareness.patch: Added for new Mesa
selinux code.
- xorg-x11-server-1.1.1-Xdmx-render-fix-fdo7482.patch: Backport a Render
fix for Xdmx.
- xorg-x11-server-1.1.1-no-composite-in-xnest.patch: Disable Composite in
Xnest, as it's known not to work.
- Enable composite by default.
- Split spiffiffity patch into one patch per change:
xorg-x11-server-1.1.0-no-move-damage.patch and
xorg-x11-server-1.1.0-dont-backfill-bg-none.patch.
- Start using the new %{dist} tag <http://fedoraproject.org/wiki/DistTag>
experimentally in the package Release field to help prevent problems like
(#197266) from occuring in the future.
